With adequate attention to the details of experiment design and the selection of participants, listening tests on loudspeakers yielded sound-quality ratings that were both reliable and repeatable. Certain listeners differed in the consistency of their ratings and in the ratings themselves. These differences correlated with both hearing threshold levels and age. Listeners with near normal hearing thresholds showed the smallest individual variations and the closest agreement with each others. Sound-quality ratings changed as a function of the hearing threshold level and age of the listener. The amount and direction of the change depended upon the specific products; some products were rated similarly by all listeners, whereas others had properties that caused them to be rated differently. Stereophonic and monophonic tests yielded similar sound-quality ratings for highly rated products, but in stereo, listeners tended to be less consistent and less critifal of products with distinctive characteristics. Assessments of stereophonic spatial and image qualities were closely related to sound-quality ratings. The relationship between these results and objective performance data is being pursued.
This paper costs $33 for non-members and is free for AES members and E-Library subscribers.